O.k. without easy and manageable tools on client side, all that is left is play, play, play and notice differences, some times subtile ones. For example, I try to focus on PU fly, trade, delivery missions, so in order to bug report in reasonable, I do restrict variance on client side interaction w/PU to a minimum.
Crashes are most of the time reported, yes. But I meant to speak about the subtile instabilities in game client interacting with the server side, like break down of keyboard input, QT targeting system, trade consoles empty ..
Some players in YT video casts say fresh servers are better – what ever that means ???
* What I noticed is that in comparison to PU 3.4.x alpha version 3.5.0 can be played in one stride c. double length of time that is c. 3-4 hours.
* By game play loops do you mean code loops like queries in relation to parcel drop off/ pick up, keyboard inputs, restricted area/or not restricted area?
* Ho do we report in break downs with out crash like: missions erased, parcel drop offs, parcel lost, no more keyboard inputs, QT mal functions, no payments, etc?
* Our efforts in the Issue Council may be lost due to that 10/10 threshold?
* How can we identify servers? Are there more than 3 (AUS,EU,US) How can we identify sub servers/instances? Can you explain the server side to us more in detail so that we know and make intelligent guesses what might be wrong (also important for error-reproduction efforts)
=> maybe a key-combination to fore a report like that crash report box with a text field while game is running and while experiencing odd server interaction could be helpful?
We host the game in the AWS cloud ( being a DevOps engineer, I’m part of the team that manages the server environments ). One of the advantages of using a cloud server provider rather than paying for our own server hardware is that when we don’t need a server anymore, we can get rid of it. Because of this, there will be points in time when there are servers that have just recently started up vs servers that have been running for a while.
To keep this brief, an example primary gameplay loop for Star Marine would be move->aim->shoot. These can also be broken down into secondary and tertiary gameplay loops but I’d recommend googling that topic because there’s some interesting articles on the subject.
One of the most important aspects of reporting a bug is knowing the steps for how it happens and this is because there can often times be multiple bugs that end up having a similar result but the thing that is breaking may be totally different. So having incredibly specific steps that allow our developers to reproduce the issue on their end helps issues get diagnosed and fixed.
A good example I can use for that is an issue we had a long time ago where ships were frequently losing their targets in Arena Commander. Most of the reports for this on the issue council weren’t specific enough with regards to how to get the issue to occur. But there was one report where a player said “get two Gladius, face each other, lock on, fly backwards, target lost”. These steps got the bug to happen every single time and that allowed the developers to figure out what was causing the issue. Additionally, fixing that problem allowed us to find that there were two other issues causing target loss in Arena Commander. So the bug for ships losing their target ended up actually being three different issues. That’s why accurate steps are really important.
The 10/10 issue council threshold has a lot of factors that go into it; chief of which is players looking at the issue, trying the steps, and commenting on it. Having good steps will really help with this but asking other players to test the issue to see if it happens for them can help as well. If they aren’t able to get it to happen on their end, then that suggests a developer might also have issues reproducing it which will make it hard to fix.
The backend for the game servers is going to change a lot over time so anything I say with regards to how it works now will not be the same in the future. With the goal of our game being to have a server mesh, the idea would be for players to basically not be able to identify what server they’re on nor when they’re transitioning from one server to another. In the event that we need to know what server someone is on to debug issues that are occurring, we do have options available to us that allow us to gather than information.